Bachelor of Science Thesis
KTH School of Industrial Engineering and Management
Energy Technology EGI-2017
SE-100 44 STOCKHOLM
Kenya meeting the electricity
demand of 2030
An assessment of how Kenya Vision 2030 and climate change
impact the optimal electricity generation mix
Anna Berggren
Sara Österberg
Bachelor of Science Thesis EGI-2017
Kenya meeting the electricity demand of
2030
Anna Berggren
Sara Österberg
Approved ExaminerMark Howells
SupervisorNandi Moksnes
KTH, SE-100 44 Stockholm. Phone: +46 8 790 6561. Fax: +46 8 790 8192. E-mail: erika2@kth.se www.kth.se/student/utlandsstudier/examensarbete/mfs
This study has been carried out within the framework of the Minor Field Studies Scholarship Programme, MFS, which is funded by the Swedish International Development Cooperation Agency, Sida.
The MFS Scholarship Programme offers Swedish university students an oppor-tunity to carry out two months’ field work, usually the student’s final degree pro-ject, in a country in Africa, Asia or Latin America. The results of the work are presented in an MFS report which is also the student’s Bachelor or Master of Sci-ence Thesis. Minor Field Studies are primarily conducted within subject areas of importance from a development perspective and in a country where Swedish in-ternational cooperation is ongoing.
The main purpose of the MFS Programme is to enhance Swedish university students’ knowledge and understanding of these countries and their problems and opportunities. MFS should provide the student with initial experience of conditions in such a country. The overall goals are to widen the Swedish human resources cadre for engagement in international development cooperation as well as to promote scientific exchange between unversities, research institutes and similar authorities as well as NGOs in developing countries and in Sweden.
The International Relations Office at KTH the Royal Institute of Technology, Stockholm, Sweden, administers the MFS Programme within engineering and applied natural sciences.
Erika Svensson Programme Officer
i
ABSTRACT
Kenya’s electricity sector faces extensive development as the country strives to achieve the goal of universal access to electricity by 2020 and the national blueprint Kenya Vision 2030. With a history of overestimating the future electricity demand, it is important to make investment decisions based on realistic demand prognoses. In this study, we evaluate a cost-optimal energy mix for electricity generation in 2030 for scenarios following business-as-usual (BAU) and Kenya Vision 2030 demand forecasts, using the spatial electrification tool OnSSET and the energy system modelling tool OSeMOSYS. Kenya is vulnerable to climate change and faces the challenge of frequent droughts, making water management a national priority. This study considers the nexus of Climate, Land, Energy and Water (CLEW). We look at how a climate change scenario, following a worst-case development according to Representative Concentration Pathway (RCP) 8.5, can affect the hydrology of the Tana River and thereby the electricity generation from its hydro power plants. Results show that the share of grid connected households increase with a higher demand forecast, that the investment cost of meeting the electricity demand of Kenya Vision 2030 is 106 % higher than the investment cost at BAU demand and that extreme climate change reduces the electricity generation of hydropower, leading up to that 2 % of electricity production must be supplied by other (non-hydro) sources in 2030. A conclusion drawn from this study is that the demand estimation has a central role for the investments in electricity technologies and is suggested to be based on trends of Kenya’s actual growth rather than visionary goals. Another conclusion is that a scenario of climate change impacts the future usage of hydropower, making the nexus to other natural resources important to include in the development of the electricity system.
SAMMANFATTNING
Kenyas elsektor står inför omfattande utveckling när landet strävar efter att ge hela befolkningen tillgång till elektricitet och att uppfylla de nationella utvecklingsmålen enligt Kenya Vision 2030. Då elbehovet tidigare överskattats är det viktigt att investeringsbeslut grundas på realistiska efterfrågeprognoser. I den här studien utvärderas en kostnadsoptimal energimix för elgenerering år 2030 i scenarion som följer prognostiserade elbehov enligt business-as-usual (BAU) och Kenya Vision
2030, med hjälp av modelleringsverktygen OnSSET och OSeMOSYS. Kenya är utsatt för
klimatförändringar och har problem med återkommande torka, därför är vattenplanering högt prioriterad i landet. Den här studien tar hänsyn till hur Klimat, Land, Energi och Vatten (CLEW) påverkar varandra. Vi tittar på hur ett extremt klimatförändringsscenario som följer Representative Concentration Pathway (RCP) 8.5 kan påverka hydrologin i floden Tana och därmed elgenereringen från dess vattenkraftverk. Resultaten visar att andelen elnätsanslutna hushåll bör bli större vid ett högre elbehov, att investeringskostnaden för att möta elbehovet enligt Kenya Vision 2030 är 106 % högre än investeringskostnaden i ett BAU-scenario och att extrema klimatförändringar kan göra att elgenereringen från vattenkraft minskar, vilket gör att 2 % av elgenereringen måste ersättas av andra energikällor år 2030. En av slutsatserna i studien är att efterfrågeprognosen har en central roll för investeringar som görs i elsystemet och bör baseras på utvecklingstrender snarare än visionära mål. En annan slutsats är att klimatförändring påverkar den framtida produktionen från vattenkraft, vilket gör att kopplingen till andra naturresurser är viktig att inkludera i utvecklingen av elsystemet.
ii
ACKNOWLEDGEMENT
We would like to acknowledge the following people for their help, advice and encouragement in the work with this Bachelor’s Thesis:
Nandi Moksnes - for your patient supervision and fast feedback. Thank you for challenging us to think for ourselves and to make decisions. We’ve learnt very much from working with you.
Professor Mark Howells - for the inspiration and for the opportunity to do this project.
Eunice Pereira Ramos, Alexandros Korkovelos and the rest of the KTH-dESA team - for introducing us to the tools for energy system modelling and for answering our questions along the way.
Maria Stridsman at the Swedish Embassy in Nairobi - for sharing knowledge about Kenya’s energy sector and for connecting us to your contacts. You helped us to get the most out of our field study. Boniface Kinyanjui at Kenya Power and Lighting Company - for deepening our understanding for Kenya’s electricity sector and for assisting us with essential data.
Dr Mary Mbithi and Dr Anthony Wambugu at University of Nairobi - for your warm welcoming at the University of Nairobi.
We would also like to thank David Mwangi, Charles Muchunku, Tom Ogol and Michelle Paula Akute for taking the time to meet us and for sharing your expertise.
We also want to express our gratitude to SIDA and ÅForsk for financing our minor field trip to Kenya. It was an inspiring experience and we have learnt a lot.
iii
TABLE OF CONTENTS
ABSTRACT ... I SAMMANFATTNING ... I ACKNOWLEDGEMENT ... II LIST OF FIGURES ... IV LIST OF TABLES ... IV ABBREVIATIONS ... V LIST OF UNITS ... V 1. INTRODUCTION ... 1 1.1BACKGROUND ... 1 1.2OBJECTIVE ... 2 1.3RESEARCH QUESTIONS... 2 1.4STUDY BOUNDARIES... 2 2. METHODOLOGY ... 3 2.1SCENARIOS ... 3 2.2CALIBRATION OF MODELS ... 4 2.2.1ONSSET ... 4 2.2.2OSEMOSYS ... 5 2.2.3WEAP ... 7 2.3MODELLING PROCESS... 8 3. RESULTS ... 93.1SPLIT OF HOUSEHOLD ELECTRIFICATION SOLUTIONS ... 9
3.2OPTIMAL ENERGY MIX FOR ELECTRICITY GENERATION TO THE NATIONAL GRID IF KENYA VISION 2030 IS ACHIEVED 10 3.3 IMPACTS ON HYDROPOWER ELECTRICITY GENERATION IN 2030 IF CLIMATE FOLLOWS REPRESENTATIVE CONCENTRATION PATHWAY 8.5 ... 11
4. DISCUSSION ... 12
5. CONCLUSIONS ... 13
6. RECOMMENDATIONS FOR FUTURE WORK ... 13
REFERENCES ... 14 APPENDIX 1 – CAPITAL COST FOR ONSSET OFF-GRID TECHNOLOGIES ... A APPENDIX 2 – ENERGY CARRIERS IN THE REFERENCE ENERGY SYSTEM ... A APPENDIX 3 – TECHNOLOGIES IN THE REFERENCE ENERGY SYSTEM ... B APPENDIX 4 – DATA COLLECTION OSEMOSYS MODEL ... E
APPENDIX 4.1–TIMESLICES OSEMOSYS ... E
APPENDIX 4.2–TECHNOLOGY COSTS ...F
iv
APPENDIX 4.4–CAPACITY FACTORS WIND ... H
APPENDIX 4.5–RESIDUAL CAPACITY ... I
APPENDIX 4.6–TOTAL ANNUALMIN CAPACITY INVESTMENT (GW) ... J
APPENDIX 4.7–SPECIFIED DEMAND PROFILE ... J
APPENDIX 4.8–TECHNOLOGY ACTIVITY ... K
APPENDIX 4.9–EMISSION ACTIVITY RATIO ... K
APPENDIX 5 – EVAPOTRANSPIRATION AND PRECIPITATION IN KENYA 2030 IN RCP 2.6 AND 8.5 ... L APPENDIX 6 – INTEGRATION OF MODELS ... M
APPENDIX 6.1–CALCULATION OF SPECIFIEDANNUALDEMAND ... M
APPENDIX 6.2–CALCULATION OF GRID PRICE ... M
APPENDIX 6.3–CALCULATION OF CAPACITY FACTORS ... M
LIST OF FIGURES
FIGURE 1 – THE CLEWS FRAMEWORK ... 2
FIGURE 2 – INTERLINKAGE OF THE MODELS ... 3
FIGURE 3 – THE OSEMOSYS TIMESLICES ... 5
FIGURE 4 – REFERENCE ENERGY SYSTEM... 6
FIGURE 5 – KENYA’S RIVER BASINS AND THE WEAP SCHEMATIC OF THE TANA RIVER BASIN (MOKSNES AND HOWELLS, 2016) ... 7
FIGURE 6 – PRECIPITATION AND EVAPOTRANSPIRATION DATA FOR RCP 2.6 AND RCP 8.5 ... 7
FIGURE 7 – THE MODELLING PROCESS ... 8
FIGURE 8 – ONSSET MAP WITH THE OPTIMAL SPLIT OF HOUSEHOLD ELECTRICITY SOLUTIONS IN THE VISION-SCENARIO ... 9
FIGURE 9 – ELECTRICITY PRODUCTION BY SOURCE IN 2030 IN VISION COMPARED TO BAU ... 10
FIGURE 10 – OSEMOSYS RESULTS FOR VISION TIMELINE OF INSTALLED CAPACITY, ELECTRICITY GENERATION BY ENERGY SOURCE AND COMPARISON TO BAU ... 10
FIGURE 11 – REDUCTION OF ELECTRICITY PRODUCTION BY HPP AND THE EFFECTS ON THE TOTAL ENERGY MIX AT VISION-DEMAND ... 11
FIGURE 12 – THE CAPACITY FACTORS FOR KAMBURU, KIAMBERE, KINDARUMA AND MASINGA HPPS IN 2030 IN RCP 2.6 AND RCP 8.5 ... 11
LIST OF TABLES
TABLE 1 – DEFINITION OF SCENARIOS: VISION, VISION – CLIMATE CHANGE AND BAU ... 4TABLE 2 – COMPARISON OF THE COST OPTIMAL SPLIT OF ELECTRICITY SOLUTIONS IN VISION AND BAU SCENARIOS ... 9
v
ABBREVIATIONS
BAU Business-as-usual
CLEW Climate, Land-use, Energy and Water
CO2 Carbon dioxide
GHG Greenhouse gases
HPP Hydro Power Plant
LCOE Levelized Cost Of Electricity
OnSSET Open Source Spatial Electrification Tool OSeMOSYS Open Source Energy Modelling System
PGTMP Power Generation and Transmission Master Plan
PP Power Plant
PV Photovoltaic
RCP Representative Concentration Pathway SDGs Sustainable Development Goals TEMBA The Electricity Model Base for Africa WEAP Water Evaluation And Planning system
LIST OF UNITS
GW Gigawatt (109 W)
GWh Gigawatt-hour (106 kWh)
hh Household
kWh Kilowatt-hour
kWh/hh/year Kilowatt-hour per household per year MtCO2eq Million tonnes of CO2-equivalents
MW Megawatt (106 W)
1
1. INTRODUCTION
1.1 BACKGROUND
Energy poverty is declared to be one of the main problems that sub-Saharan Africa is facing. In 2014, more than 620 million people in sub-Saharan Africa didn’t have access to electricity in their homes (International Energy Agency 2014). “To ensure access to affordable, reliable, sustainable and modern energy for all” is the 7th (UNDESA 2016) of The Sustainable Development Goals (SDGs), agreed upon
by world leaders in 2015 as a part of The 2030 Agenda – a global action plan for people, planet and prosperity (United Nations 2016). The Government of Kenya has declared the target to reach universal access to electricity by 2020 (The Government of Kenya 2007). In 2015, the percentage of electrified households was 46 % (Power Africa 2016), compared to 23 % in 2012 (The World Bank 2012). The rate of new grid connections is high, partly thanks to the rural electrification programme “The Last Mile Connectivity Project” (African Development Bank Group 2017), carried out by The African Development Bank, The Government of Kenya and The World Bank to supply households with electricity from on- and off-grid electricity solutions (Ministry of Energy 2016).
Increasing access to energy has positive effects on other SDGs since it is closely linked to economic and social development (UNDESA 2016). Achieving the goal of universal access to electricity is crucial to carry out Kenya’s long-term national planning strategy, Kenya Vision 2030, containing ambitious goals for economic, social and political development. The main vision is to transform Kenya into “a newly industrializing middle-income country providing a high-quality life to all its citizens by the year 2030” (The Government of Kenya 2007). Kenya’s population is estimated to increase from 46 million to 65.4 million in the years 2015-2030 (The World Bank Group 2017a) and the electricity consumption is forecasted to increase with 315 % during that period in case Kenya Vision 2030 and universal access to electricity are achieved (International Energy Agency 2017). This implies that Kenya’s electricity sector must develop to meet a much higher demand and there are several flagship projects for electricity generation planned (Ministry of Energy and Petroleum 2016).
Historically, hydropower has been the most used source for electricity generation in Kenya (IEA 2017). Today, the reliability of hydropower is questioned (Wahome 2013) as periods of droughts have affected the output of hydropower plants (HPPs) the last years (Ministry of Environment and Natural Resources 2016). Climate change impacts the patterns of rainfall and temperature, causing periods of floods and droughts (Ministry of Environment and Mineral Resources 2013). Kenya has a high vulnerability to these fluctuations in water availability since it destroys livelihoods and causes hunger (Kenya National Bureau of Statistics 2014), disease and even death (National Drought Management Authority 2013). Access to clean drinking water (SDG6) and security in food supplies (SDG2)cannot be endangered when expanding the energy system. Water can’t be replaced when it comes to drinking and irrigation. Meanwhile, Kenya has a good potential to use energy from wind and sun as well as geothermal energy located in the Rift Valley (Energy Regulatory Commission 2012) for future electricity generation to replace hydropower if that is necessary.
2
In this study the Climate, Land-use, Energy and Water strategies (CLEWs) framework has been applied to model the nexus of these natural resources in Kenya (Figure 1). The idea behind CLEWs is that the natural resources of land, energy and water are highly integrated and that the exploitation of them contributes to climate change, meanwhile they are all highly vulnerable to it. The interlinkages between those systems must be considered to develop consistent strategies and ensure an efficient resource management (Bazilian et al. 2011; Howells et al. 2013). As a part of our work with this Bachelor’s thesis, we have conducted a field study in Nairobi. We met professionals in Kenya’s energy sector1 to
ask for their opinions on the main challenges for Kenya’s electricity generation. Some of them expressed concern about the risk of overestimating the future demand, since previous demand forecasts for Kenya has regularly been too optimistic (Ministry of Energy and Petroleum 2016). Kenya has the goal to sustain a yearly GDP growth of 10 % (The Government of Kenya 2007) to achieve Kenya
Vision 2030. There is a risk of overestimating the need of capacity investments if demand-prognoses
are based on a correlation with the 10 % GDP growth (Power Africa), since this is a very ambitious goal compared to earlier GDP growth in Kenya (varying in the range 0.2 % to 8.4 % in the period 2001-2014 (The World Bank Group 2014)). The World Bank projects the GDP growth in Kenya to be 5.8 % in 2018 and 6.1 % in 2019 (The World Bank Group and Kenya Economic Roundtable 2017).
1.2 OBJECTIVE
The objective of this study is to investigate how the choice of demand estimation for 2030 and the climate change affects the energy mix for electricity generation in Kenya.
1.3 RESEARCH QUESTIONS
1. How does the size of electricity demand forecasted for Kenya Vision 2030 change the optimal split of household electrification solutions based on spatial electrification analysis, compared to a BAU demand?
2. How does the size of electricity demand forecasted for Kenya Vision 2030 change the cost-optimal energy mix for on-grid electricity generation compared to a BAU demand?
3. What impact will climate change according to RCP 8.52 have on the electricity generation from
hydropower in Tana River by 2030 and what are the effects on the on-grid electricity generation mix?
1.4 STUDY BOUNDARIES
• Kenya’s electricity system is modelled. No electricity trade with neighbouring countries has been included.
• This study investigates a simplification of reality where only households are given the possibility to get electricity from off-grid solutions.
• The optimization in this study runs from year 2015 till year 2030.
• The impacts of climate change have only been modelled for the HPPs in the Tana river basin.
1 The professionals that we have talked to about the development of Kenya’s electricity generation is:
Maria Stridsman - Swedish Embassy in Nairobi
Boniface Kinyanjui – Chief Engineer, Generation Planning at Kenya Power and Lightning Company David Mwangi – Senior Energy Consultant
Charles Muchunku – Independent Renewable Energy Consultant
2 RCP is short for Representative Concentration Pathway. The scenario RCP 8.5 is a worst-case scenario of climate change
where the concentration of greenhouse gases don’t stagnate within the 21th Century (Riahi, Grübler, and Nakicenovic 2007).
Climate Energy Water and Land FIGURE 1–THE CLEWS FRAMEWORK
3
2. METHODOLOGY
An electricity demand for development in accordance with Kenya Vision 2030 (Vision) was compared to a lower demand based on development following historical growth, referred to as BAU. Two extremes of climate change scenarios, RCP 2.6 (van Vuuren et al. 2007) and RCP 8.5 (Riahi, Grübler, and Nakicenovic 2007), were applied to the scenario with Vision-demand.
Three modelling softwares were used to investigate what mix of electricity generation sources Kenya should aim to have in 2030, taking into account the nexus of Climate, Land, Energy and Water (Howells et al. 2013; Bazilian et al. 2011) (Figure 2). Energy was modelled in the Open Source Spatial Electrification Tool (OnSSET) (Mentis et al. 2015) and in The Open Source Energy Modeling System (OSeMOSYS) (Howells et al. 2011). Water and Land-use were modelled in the Water Evaluation and Planning tool (WEAP) (Sieber 2017), using an existing model of the Tana catchment (Moksnes and Howells 2016). To include Climate in the nexus assessment, this model was extended with RCP-specific precipitation and evapotranspiration data to see how climate change affects the hydrology in the Tana river, and thereby the capacity factors of its hydropower plants in 2015-2030.
2.1 SCENARIOS
If Kenya Vision 2030 is achieved, the national electricity demand is expected to increase rapidly, starting at 7655 GWh in 2015 (IEA 2017). In the Power Generation and Transmission Master Plan (PGTMP) a similar Vision-scenario is investigated and they project a demand of 39 260 GWh in 2030 (Ministry of Energy and Petroleum 2016). That demand prognosis is used in the Vision scenario. The average electricity consumption per household and year was set according to Tier 3 (695 kWh/hh/year) following Nerini et al. (Nerini et al. 2016). This level of electricity consumption is enough to power a washing machine, computer and lights – a set of household appliances that can be argued necessary to provide all citizens ”a high-quality life” (Rosling 2010), which Kenya aims to achieve by 2030 (The Government of Kenya 2007). A BAU scenario with a lower level of national electricity demand was created to put the Vision scenario in a context. The electricity demand in BAU was set to 27 366 GWh (collected from the reference-scenario in PGTMP) (Ministry of Energy and Petroleum 2016), following historical trends of growth in Kenya. The household electricity consumption was set to 484 kWh/hh/year, assuming that residential electricity demand in BAU represents the same share of the total national electricity demand as in Vision. Both scenarios are based on that universal access to electricity is achieved in Kenya before 2030.
To understand the impact of climate change, two different scenarios of future climate were studied. The smallest possible climate change was used as a reference with a projected peak of global emissions of greenhouse gases (GHGs) before year 2020, following the Representative Concentration Pathways 2.6 (van Vuuren et al. 2007). RCP 2.6 implies an increase of global mean temperature limited to two degrees Celsius (Symon 2013) and is in line with trends of flat emissions of greenhouse gases the past
OnSSET
OSeMOSYS
WEAP
Grid cost of electricity Grid electricity demand Demand fromhydro power plants Constraints for
hydropower
CO2 emissons
Temperature Rain
4
years (International Energy Agency 2016). This scenario was compared to a scenario of extreme climate change with climate data from RCP 8.5 where the concentration of GHGs is not expected to stagnate within the 21th Century (Riahi, Grübler, and Nakicenovic 2007).
The demand levels were combined with climate predictions into three research scenarios: Vision,
Vision – Climate Change and BAU. Their characteristics are presented in Table 1. TABLE 1–DEFINITION OF SCENARIOS:VISION,VISION –CLIMATE CHANGE AND BAU
SCENARIO: VISION VISION - CLIMATE CHANGE BAU NATIONAL ELECTRICITY DEMAND 2030 Vision demand 39 260 GWh Vision demand 39 260 GWh BAU demand 27 366 GWh HOUSEHOLD ELECTRICITY CONSUMPTION 2030 695 kWh 695 kWh 484 kWh CLIMATE SCENARIO RCP 2.6 RCP 8.5 RCP 2.6
2.2 CALIBRATION OF MODELS
2.2.1
O
NSSET
The Open Source Spatial Electrification Tool (OnSSET) is developed by KTH Royal Institute of Technology in collaboration with UNDESA as a part of reaching the goal of universal access to electricity globally (Mentis et al. 2015). It calculates an optimal split of on-grid, mini grid and standalone electrification solutions for unelectrified households to reach a target level of electricity consumption, comparing the Levelized Cost Of Electricity (LCOE) (Nerini et al. 2016). OnSSET takes local characteristics such as population density, distance to transmission lines and available energy sources into account, and investments are made in a single timestep over the whole modelling period. The cost optimization tool OnSSET (Mentis et al. 2015) was used to calculate the split of household electrification technologies and the electricity demand on the national grid in the scenarios of Vision and BAU demand. The household electrification technologies included in this optimization were: national grid; mini grids of diesel, hydro, solar photovoltaic (PV) and wind; standalone diesel and solar PV. The modelling period in OnSSET was 2015-2030. At the start of this period, 46 % of the households had access to on-grid electricity (USAID 2017), and by 2030, universal access shall be achieved according to SDG7 (UNDESA 2016). In 2015-2030 the population will grow from 46.1 million to 65.4 million (The World Bank Group 2017a), the urban ratio increase from 25.6 % to 32.97 % (The World Bank Group 2017b) and the number of people per household is 4.4 (Michael Bauer Research 2015). For the national grid a capacity investment cost of 1935 USD/kW (KTH-dESA 2017), an existing grid-cost ratio of 0.1 (KTH-dESA 2017), a base-to-peak ratio of 0.38 (KPLC 2016) and losses at an average of 14.75 % (KPLC 2017) were used. An initial grid price of 0.052 USD/kWh was assumed before calculating scenario-specific grid price from OSeMOSYS. The diesel price used was 0.760928 USD/liter, calculated as an average over the modelling period (Ministry of Energy and Petroleum 2016). The capital costs for the off-grid electrification technologies used are presented in Appendix 1.
5
2.2.2
OS
EMOSYS
The Open Source Energy Modelling System (OSeMOSYS) is a tool for long-run energy planning (Howells et al. 2011). It is a flexible multi-year tool that calculates the lowest net present cost of an energy system to meet a specified demand within given constraints. The open source basis of the tool makes it accessible to students, business analysts, government specialists and developing country energy researches which aims to create a better understanding of energy modelling in the developing countries (Howells et al. 2011).
OSeMOSYS was used to optimize the energy mix for the national grid to meet Kenya’s 2030 electricity demand in the different scenarios. Kenya-specific data from The Electricity Model Base for Africa (Taliotis et al. 2016) was used as a foundation for sets and parameters in OSeMOSYS. The cost optimization was made with a discount rate of 5 %. The modelling period was set to start in 2015, for the possibility to use historical data to calibrate the base year, and end in 2040 to avoid unwanted edge effect around 2030. The model was divided into 36 Timeslices. The days were divided into three Timeslices (see Figure 3) to capture the fluctuations in electricity demand and the hours of sunshine. Over the year the 12 calendar months were used as Timeslices, to capture the variations in capacity factors of wind (extracted using the Renewable Ninja Tool) (see Appendix 4.4) (Staffell and Pfenninger 2016; Pfenninger and Staffell 2016) and hydropower modelled in WEAP. The existing HPPs Masinga, Gitaru, Kiambere, Kamburu and Kindaruma and the future HPP candidate High Grand Falls were modelled separately in OSeMOSYS with all their associated parameters, since these technologies were modelled in WEAP. Nuclear has been included as a technology since it is a candidate for future generation expansion, introduced earliest in 2028 (Ministry of Energy and Petroleum 2016). A representation of technologies who produce, use & convert, and transfer energy carriers is found in the Reference Energy System (Figure 4). The
ResidualCapacity was set in accordance with statistics from the Energy Regulation Commission (Energy Regulatory Commission 2015). TotalMinAnnualCapacityInvestment represents actual investments in power plants for the years 2016-2020, presented in PGTMP (Ministry of Energy and Petroleum 2016), since these projects has reached financial closure at the date of this report. TotalAnnualMaxCapacityInvestment has been set to 100 MW for all technologies in the years 2018-2020, except for geothermal power plant since capacity of 318.5 MW is planned in 2019 (Ministry of Energy and Petroleum 2016), to disable OSeMOSYS from investing in large projects that are not scheduled in reality. Other techno-economic parameters such as costs associated with investment and operation, availability factors, capacity factors, efficiency and lifetime of technologies were collected from various sources (see Appendix 4).
6
All costs associated with transmission and distribution technologies were set to zero since they are modelled in OnSSET, not to be accounted for twice. The grid electricity losses were set to 17.5 % in 2015 and assumed to decrease towards KPLC’s goal of 12 % losses in 2030 (KPLC 2017). ReserveMargin was assumed to increase from 1 % in 2015 to 18 % in 2030, even though this reserve margin does not correctly reflect the reality of 2015. This assumption was made due to lack of data of the actual reserve margin in 2015 and to reflect a situation of frequent power cuts. Kenya has the intention to limit its total CO2-emissions 30 % below a business as usual scenario of 143 MtCO2eq per year by 2030 (Ministry
of Environment and Natural Resorces 2015). Based on emission data from earlier years, the electricity generation accounts for approximately 22% of the total emissions (IndexMundi 2017). With the assumption that this emission ratio stays constant and all sectors stay 30 % below their business as usual projection, the upper limit for CO2 emissions was set to 22 MtCO2eq for 2015-2030. After 2030
they could increase at the same rate as the electricity demand, since this is beyond Kenya’s long-term plans.
FIGURE 4–REFERENCE ENERGY SYSTEM
7
2.2.3
WEAP
WEAP (Water Evaluation And Planning) is a software for integrated water resources planning, developed by Stockholm Environmental Institute. It operates on the principle of a water balance and can be applied to simulate a broad range of natural and engineered components in different kinds of water systems, for example a river basin, over a given period. A WEAP-model can be customized to meet the requirements of the analysis and data availability of the system (Sieber 2017).
WEAP was used to simulate the future potential of hydropower in the climate scenarios. An existing WEAP model from UN Sustainable
development goals from a Climate Land Energy and Water perspective for Kenya by Moksnes (2016) was
used and extended with data for climate change. The model represents the Tana basin, supplying major irrigation sites with water
(Moksnes and Howells 2016) and containing 599
MW of Kenya’s 820 MW installed hydropower capacity (Energy Regulatory Commission 2015). The Tana basin is interesting for a CLEWs analysis since it is a system that uses Water for agriculture (Land-use) and hydropower (Energy).
The Moksnes-model was adjusted to simulate the hydrology in the river in the scenarios of climate change (RCP 2.6 and RCP 8.5), using a simplified method. The focus was to update the climate scenario-specific parameters that have most impact on the hydrology of the river - monthly precipitation and monthly evapotranspiration for all irrigation sites of the basin. Precipitation-data was extracted for the climate scenarios and used as input without adjustments, see Figure 6. Data for variations in relative humidity and mean, max & min temperatures were extracted and recalculated into monthly evapotranspiration data, see Figure 6, using the ETo Calculator version 3.2 (Food and Agriculture Organization of the United Nations 2017) developed by the Food and Agriculture Organization of the United Nations. All climate data was extracted monthly over the modelling period 2012-2040 from the CMIP5 multi model ensemble supplied by World Climate Research Programme's Working Group on Coupled Modelling (Taylor, Stouffer, and Meehl 2012), available at KNMI Climate Explorer (The Royal Netherlands Meteorological Institute (KNMI) 2017). Data was extracted for the coordinates 35-40 E, 2.5-0 S which represents the area of the Tana basin. The hydropower priority was set to 99 (lowest possible) for all reservoirs in the model since we aimed to reflect a situation where clean water for the cities and for irrigation in the agriculture areas isn’t endangered. Due to lack of data the model was not recalibrated after changing values for evapotranspiration and precipitation in this work.
FIGURE 5–KENYA’S RIVER BASINS AND THE
WEAP SCHEMATIC OF THE TANA RIVER BASIN
(MOKSNES AND HOWELLS,2016)
8
2.3 MODELLING PROCESS
Following process was carried out when extracting results for the three scenarios:
1. An OnSSET optimization was performed using the value 0.052 USD/kWh as an initial price for grid electricity. This resulted in a first on- vs off-grid electrification split.
2. The demand parameters for OSeMOSYS (SpecifiedAnnualDemand) were calculated (see Appendix 6.1). A constant growth-rate was assumed for the demand from 2015 to 2030, continuing throughout the modelling period. SpecifiedAnnualDemand parameters were imported to OSeMOSYS.
3. A first optimization in OSeMOSYS (part 1) was performed with CapacityFactors for the HPPs modelled in WEAP set to 100 %. This resulted in a cost optimal generation from these HPPs, as if they were not restricted by variations in Capacity Factors within the years.
4. The OSeMOSYS-result ProductionByTechnology was extracted and used as Demand-input for the HPPs in WEAP.
5. The WEAP model predicted how the water availability allowed the power plants to meet the demand during each month of the modelling period.
6. From the results of Hydropower Generation in WEAP, the CapacityFactors for the HPPs were calculated (see Appendix 6.3).
7. The WEAP-CapacityFactors were inputted to OSeMOSYS for Masinga, Kindaruma, Kiambere and Kamburu HPPs. Noticing that the results for Gitaru HPP were not realistic, the CapacityFactors for Kiambere were used since its installed capacity is of similar size3. The
categoric HPP-technologies in OSeMOSYS were given CapacityFactors of Kiambere (Large) and Kamburu (Medium) since we wanted to represent the variation in hydrology for all of Kenya’s HPPs. After these adjustments, OSeMOSYS calculated a new cost-optimization.
8. With the results from OSeMOSYS part 2, a new grid price was calculated (see Appendix 6.2). 9. The new price of grid electricity was used as input in OnSSET part 2, resulting in an updated
version of the optimal split of on- and off-grid household electrification solutions.
10. The SpecifiedAnnualDemand for residential and industrial on-grid electricity was calculated (see Appendix 6.1).
11.
With the new SpecifiedAnnualDemand, the final results were extracted in OSeMOSYS part 3.
3Gitaru HPP has a capacity of 225 MW and Kiambere 168 MW (Energy Regulatory Commission 2015).
FIGURE 7–THE MODELLING PROCESS Extraction of specific demand for WEAP HPPs
Calculation of CapacityFactors for WEAP HPPs
Calculation of new price for grid electricity Calculation of grid electricity demand Calculation of grid electricity demand OnSSET part 1 OSeMOSYS part 1 WEAP OSeMOSYS part 2 OnSSET part 2 OSeMOSYS part 3 Step 1. Step 2. Step 3. Step 4. Step 5. Step 6. Step 7. Step 8. Step 9. Step 10. Step 11.
9
3. RESULTS
3.1 SPLIT OF HOUSEHOLD ELECTRIFICATION SOLUTIONS
TABLE 2–COMPARISON OF THE COST OPTIMAL SPLIT OF ELECTRICITY SOLUTIONS IN VISION AND BAU SCENARIOS
The least cost split of electrification solutions shows that the areas where Kenya’s major cities are located should be connected to the national grid, while rural areas with a lower population density should be electrified with off-grid solutions. The comparison of the split of electrification solutions in the Vision and BAU scenarios shows that the higher household consumption in Vision results in a larger share of the population connected to the national grid. The most common off-grid solution at both demand levels is standalone photovoltaic. At the BAU demand, only a little hydro and no wind is installed as mini grid technologies. When the demand increases to the Vision-level, mini grid hydro increases its share and mini grid wind is introduced. Even though they support more people with electricity, these technologies are not as visual on the map as standalone diesel. This can be explained partly by the fact that there needs to be a larger population in a settlement for a mini grid to be more cost-effective than a standalone. The use of diesel as a source of electricity is slightly lower in Vision, reflecting the situation of a low investment cost but a high variable cost (diesel fuel) that makes diesel a less suitable alternative at high consumption levels.
Scenario: Vision BAU
National grid 93,18% 90,37% Standalone diesel 0,009% 0,014% Standalone PV 6,53% 9,57%
Mini grid diesel 0% 0%
Mini grid PV 0% 0%
Mini grid wind 0,11% 0%
Mini grid hydro 0,17% 0,05%
FIGURE 8 – ONSSET MAP WITH THE OPTIMAL SPLIT OF HOUSEHOLD ELECTRICITY SOLUTIONS IN THE VISION-SCENARIO
10
3.2 OPTIMAL ENERGY MIX FOR ELECTRICITY GENERATION TO THE NATIONAL GRID
IF KENYA VISION 2030 IS ACHIEVED
The results of the OSeMOSYS optimization is presented in Figure 9 and Figure 10. Capacity installations follow TotalMinAnnualCapacityInvestment (see 2.2.2 OSeMOSYS) during the first five years and the only deviations from the plans are small increases in hydropower, geothermal and wind. We notice a switch in the capacity installations around 2023. After this year, no additional capacity of hydro, wind and coal is installed. Instead solar PV (2022) and natural gas (2024) are introduced to the energy mix and increase in capacity throughout the modelling period. No oil is installed during the modelling period and its residual capacity is phased out and replaced with other fossil fuel candidates. The optimization suggests that electricity demand shall be covered without installation of nuclear. A comparison of the installed capacity in the scenarios Vision and BAU shows that solar PV, natural gas, geothermal, coal and diesel are expanded to meet the higher demand. Diesel and geothermal increase their shares in the mix substantially (with 6.7 % respectively 3.9 %), which results in that the shares of especially hydro and wind decrease. The total discounted investment cost for the capacity installed in Vision is 38400 Million USD which is 106 % higher than BAU.
Geothermal is the main source of electricity in the production mix suggested for Vision, followed by hydro and solar PVs. The production mix is dominated by CO2
-neutral sources as electricity from fossil fuels has a low contribution to the mix (only 5 %). If looking at the graphs of installed capacity and electricity generation it is noticeable that 2.30 GW of diesel capacity is installed in 2030, but it does not contribute to the production at all – this indicates that diesel is installed as reserve capacity.
FIGURE 10–OSEMOSYS RESULTS FOR VISION
TIMELINE OF INSTALLED CAPACITY,ELECTRICITY GENERATION BY ENERGY SOURCE AND COMPARISON TO BAU FIGURE 9–ELECTRICITY PRODUCTION BY SOURCE
11
3.3 IMPACTS ON HYDROPOWER ELECTRICITY GENERATION IN 2030 IF CLIMATE
FOLLOWS REPRESENTATIVE CONCENTRATION PATHWAY 8.5
Figure 12 shows how the capacity factors vary over the year, with a peak in December-January, after a period of rain in October-November, and a common low in August-October following a dryer period in June-September (see Appendix 5). The capacity factor curves in Vision - Climate Change show bigger fluctuations within the year and lower averages, compared to Vision. This applies to all four HPPs modelled.
The HPPs generate less electricity in the Vision - Climate Change scenario compared to Vision, because of the lower capacity factors. The total reduction in electricity production from the HPPs (including the HPPs that were not modelled in WEAP (2.3 Modelling process, step 6)) is 198 GWh in 2030. Figure 11 shows how the energy mix is affected by this reduction of hydropower electricity generation. Following this change, the electricity production from geothermal energy decreases, resulting in a total generation cut of 350 GWh from these two sources combined. It is mainly covered by solar PV, but natural gas and coal increase in production as well.
To summarize the changes in the electricity generation mix for the Vision - Climate Change scenario compared to the Vision scenario: biomass and wind remain unchanged; coal, natural gas and solar PV increase (to 1.78 %, 3.09 % and 10.50 %) while hydro, oil and geothermal decrease (to 17.6 %, 0.02 % and 62.2 %). The additional capacity investment cost to fill the on-grid demand is 182 MUSD in Vision
- Climate Change compared to Vision.
FIGURE 12–THE CAPACITY FACTORS FOR KAMBURU,KIAMBERE, KINDARUMA AND MASINGA HPPS IN 2030 IN RCP2.6 AND RCP8.5
FIGURE 11–REDUCTION OF ELECTRICITY PRODUCTION BY HPP AND THE EFFECTS ON THE TOTAL ENERGY MIX
12
4. DISCUSSION
The cost optimization suggests that 93.2 % of the population should be connected to the national grid in 2030 in Vision. When comparing Vision to BAU, we see that a higher household electricity demand increases the share of households connected to the national grid. This is also suggested by Nerini et al. (Nerini et al. 2016), explaining that the high capital cost of grid connection becomes more economic on a cost per kilowatt basis at higher demand and high population density. To electrify Kenya in the most financially effective way by 2030, the forecast of future household electricity consumption – on which investment decisions are made – will play an important role.
The earlier mentioned Last Mile Connectivity project was initiated as one step towards reaching the national goal of universal access to electricity by 2020 (African Development Bank Group 2017), by connecting new consumers to the national grid and thereby enabling them to use electricity. Initially, these households will use low-Tier. In an assessment of universal access to electricity in Kenya by 2020, Power Africa (Power Africa 2016) estimates that 70-80 % should be connected to the national grid at a consumption level of Tier-1 and Tier-2. This indicated that off-grid solutions are better suited for low-electricity demand, while a higher demand makes connection to the grid more cost-effective. The investment horizon plays an important role, since the short-term national goal of universal access to electricity by 2020 (low-Tier) and the longer-term Kenya Vision 2030, to give Kenya’s citizens a high-quality life (indicating at least Tier-3), will suggest different cost-optimal splits of electrification technologies. If both goals are to be achieved, the best solution might be for rural households to gain access to electricity by 2020 with off-grid solutions. In the long-run this early gained access will contribute to reaching higher Tiers and thereby justify the investment of connecting additional households to the national grid.
As discussed in the previous section, the size of the demand plays an important role in cost-optimizations of the electricity system. In OSeMOSYS, a higher demand gives a multiplier effect on the size of installed capacity needed to meet peak demand and reserve margin. Our results show that size of demand corresponding to BAU requires 8.48 GW of installed capacity, while demand corresponding to Vision requires a 48.8 % higher capacity at 12.6 GW. At the same time, the total discounted investment cost is 106 % higher in Vision than BAU, indicating a leverage effect on the investment costs required to install the bigger capacity. This result can be explained by that the least-cost power plants are utilized to meet BAU demand and when the demand increases in Vision more cost intensive technologies are installed. Even if our results have many uncertainties (for example in economic input data, reserve capacity and maximum limits for possible installations) this quantifies how an overestimation of the demand can potentially lead to additional costs of installing excess capacity. That can happen in Kenya since projects for electricity generation are scheduled to meet the demand of Kenya Vision 2030 (Ministry of Energy and Petroleum 2016), while it is not definite that the vision will be achieved. One of the flagship projects in PGTMP is Lamu Coal PP (1000 MW), planned to be built in 2021-2023 in the coastal region near Kenya’s second largest city Mombasa (Ministry of Energy and Petroleum 2016). In our optimization, a total installed capacity of 570 MW coal is suggested in 2030 in Vision, indicating that the Lamu Coal PP might not be cost optimal at that size. Many of the people in Kenya’s energy sector that we have met for discussions have indicated that there is an imbalance in Kenya caused by that demand and production are located in different areas. This leads us to one of the limitations of the OSeMOSYS optimization: it does not include a spatial analysis which would be important to consider, especially as fuel prices differ within the country and the losses in transmission are high.
Comparing the results of Vision - Climate Change to Vision gives an insight to the possible restrictions on the use of hydropower, caused by climate change according to RCP 8.5. The capacity factors in
13
Vision - Climate change show more variability and lower average values for all HPPs that were
modelled, resulting in a reduced electricity generation from the same size of installed capacity of HPPs. An installation of 0.08 GW additional capacity of solar PV, coal and natural gas is needed to cover the production loss. This indicates that the availability of water for the HHPs changes the optimal electricity generation investments and needs to be considered when planning for future power plants. It is hard to come up with a more extended analysis for how climate change will impact the use of hydropower in Kenya, since the results presented in this study are based on the short modelling period of 15 years and does only investigate the most extreme of the climate change RCPs. The WEAP-modelling was furthermore built on the assumptions that it would be possible to prioritize water for cities and irrigation higher than water to the HPPs. This prioritization has had an impact of the results in each scenario, although it might not be fully realistic.
5. CONCLUSIONS
This study shows that the estimations of future electricity demand play a central role when making investment decisions for electrification technologies and when planning a cost optimal electricity generation mix for the national grid. It is suggested that the forecasted demand-level is based on the trends of Kenya’s actual growth rather than visionary goals. The study shows that climate change will impact the reliance of hydropower which proves the importance to understand the connection of climate, land, energy and water when developing an electricity system that meets Kenya’s future demand.
6. RECOMMENDATIONS FOR FUTURE WORK
o Trade with neighbouring countries should be included to not over-estimate the need of electricity generation in Kenya, since cross-border electricity connections can lead to less long-term investments and fuel costs in Africa (Pappis 2016).
o The on-grid optimization could be improved by including a spatial analysis since that would capture the distances of transmission and make sure that the electricity generation and demand are balanced over the map.
o Since geothermal, coal, natural gas and oil depend on water availability (Kęsicki et al. 2016) they should be included in the climate change analysis to get a complete picture of the effects on the energy system. The modelling period of climate change scenarios should be extended to reflect the slow changes and long-term effects.
o
Since Kenya’s economic development is highly dependent on climate sensitive sectors (Ministry of Environment and Mineral Resources 2013) and the future demand has correlations with GDP-growth (Power Africa 2016), it would be interesting to include research about how extreme weather caused by climate change affects GDP and the electricity demand.o
Since we have concluded that the forecasted electricity demand is important for the investment decisions, we suggest an extensive demand-analysis to be included in future work. It should take the correlation of social development with electricity demand growth, traditions that may impact electricity consumption and energy efficiency improvements into consideration.14
REFERENCES
African Development Bank Group. 2017. “Kenya - Last Mile Connectivity Project.” African Development
Bank. https://www.afdb.org/en/projects-and-operations/project-portfolio/p-ke-fa0-010/.
Bazilian, Morgan, Holger Rogner, Mark Howells, Sebastian Hermann, Douglas Arent, Dolf Gielen, Pasquale Steduto, et al. 2011. “Considering the Energy, Water and Food Nexus: Towards an Integrated Modelling Approach.” Energy Policy 39 (12): 7896–7906. doi:10.1016/j.enpol.2011.09.039.
Energy Regulatory Commission. 2012. “Renewable - Geothermal.” Renewable Energy Portal. http://www.renewableenergy.go.ke/index.php/content/28.
———. 2015. “ERC Annual Report & Financial Statements 2014/2015.” Annual Report. Nairobi: Energy
Regulatory Commission.
http://erc.go.ke/index.php?option=com_docman&task=cat_view&gid=39&Itemid=429. Food and Agriculture Organization of the United Nations. 2017. “ETo Calculator.” May 12.
http://www.fao.org/land-water/databases-and-software/eto-calculator/en/.
Howells, Mark, Sebastian Hermann, Manuel Welsch, Morgan Bazilian, Rebecka Segerström, Thomas Alfstad, Dolf Gielen, et al. 2013. “Integrated Analysis of Climate Change, Land-Use, Energy and Water Strategies.” Nature Climate Change 3 (7): 621–26. doi:10.1038/nclimate1789.
Howells, Mark, Holger Rogner, Neil Strachan, Charles Heaps, Hillard Huntington, Socrates Kypreos, Alison Hughes, et al. 2011. “OSeMOSYS: The Open Source Energy Modeling System: An Introduction to Its Ethos, Structure and Development.” Energy Policy, Sustainability of biofuels, 39 (10): 5850–70. doi:10.1016/j.enpol.2011.06.033.
IEA. 2017. “IEA Statistics - Kenya: Electricity and Heat.” International Energy Agency. https://www.iea.org/statistics/statisticssearch/report/?country=KENYA&product=electricitya ndheat&year=2010.
IndexMundi. 2017. “Kenya - CO2 Emissions.” Data Portal. IndexMundi.
https://www.indexmundi.com/facts/kenya/co2-emissions.
International Energy Agency. 2014. “Africa Energy Outlook.” World Energy Outlook 2014 Special Report. Paris: OECD/IEA. http://www.iea.org/publications/freepublications/publication/weo-2014-special-report-africa-energy-outlook.html.
———. 2016. “Key CO2 Emissions Trends.” International Energy Agency.
KenGen. 2015. “KenGen Integrated Annual Report & Financial Statements 2015.”
http://www.kengen.co.ke/sites/default/files/financial-reports/2015%20KenGen%20Annual%20Report.pdf.
Kenya National Bureau of Statistics. 2014. “Kenya Food Deficit - Kenya Data Portal.” Knoema. http://kenya.opendataforafrica.org//hiwkfyg/kenya-food-deficit.
Kęsicki, Fabian, Molly A. Walton, Tim Gould, Laura Cozzi, Kamel Ben Naceur, Paul Hugues, Raykyung Park, et al. 2016. “Water Energy Nexus.” Exerpt from the World Energy Outlook 2016. Paris: OECD/IEA.
KPLC. 2016. “KPLC Load Halfhourly 2015 (Provided for This Study after Request).” ———. 2017. “KPLC Study Data (Provided for This Study after Request).”
KTH-dESA. 2017. “Integration between ONSSET and OSeMOSYS (via MoManI).” KTH-dESA. Lako, Paul. 2010. “Technology Brief E06 - May 2010.” International Energy Agency ETSAP.
Mentis, Dimitrios, Manuel Welsch, Francesco Fuso Nerini, Oliver Broad, Mark Howells, Morgan Bazilian, and Holger Rogner. 2015. “A GIS-Based Approach for Electrification planning—A Case Study on Nigeria.” Energy for Sustainable Development 29: 142–150. doi:10.1016/j.esd.2015.09.007.
Michael Bauer Research. 2015. “Kenya Average Household Size.” ArcGIS.
15
Ministry of Energy. 2016. “Last Mile Connectivity Project.” Republic of Kenya. Ministry of Energy and
Petroleum. http://www.energy.go.ke/index.php/projects/245-last-mile-connectivity-project.html.
Ministry of Energy and Petroleum. 2016. “Development of a Power Generation and Transmission Master Plan, Kenya Long Term Plan 2015 - 2035 – Vol. I.” Bad Vilbel, Germany: Ministry of
Energy and Petroleum.
http://www.erc.go.ke/index.php?option=com_content&view=article&id=167&Itemid=680. Ministry of Environment and Mineral Resources. 2013. “Kenya’s National Climate Change Action Plan.”
Government of Kenya. https://cdkn.org/wp-content/uploads/2013/03/Kenya-National-Climate-Change-Action-Plan.pdf.
Ministry of Environment and Natural Resorces. 2015. “Kenya’s Intended Nationally Determined Contribution (INDC).” Ministry of Environment and Natural Resources. http://www.environment.go.ke/wp-content/uploads/2015/07/Kenya_INDC_20150723.pdf. Ministry of Environment and Natural Resources. 2016. “Kenya National Adaptation Plan 2015-2030.”
Government of Kenya.
http://www4.unfccc.int/Nap/Documents%20nap/Kenya_Nap_Final.Pdf.
Moksnes, Nandi, and Mark Howells. 2016. “UN Sustainable Development Goals from a Climate Land Energy and Water Perspective for Kenya.” KTH Industrial Engineering and Management. http://kth-primo.hosted.exlibrisgroup.com/primo_library/libweb/action/search.do?ct=search&fn=searc h&initialSearch=true&mode=Basic&tab=default_tab&indx=1&dum=true&srt=rank&vid=46KT H_VU1&frbg=&prefLang=sv_SE&scp.scps=scope%3A%2846KTH_DATABASER_DS%29%2Csco pe%3A%2846KTH_ALMA_DS%29%2Cscope%3A%2846KTH_DIVA_DS%29%2Cprimo_central_ multiple_fe&vl%28freeText0%29=nandi+moksnes.
National Drought Management Authority. 2013. “Drought Risk Management and Ending Drought Emergencies.” Republic of Kenya. http://www.ndma.go.ke/resource-center/ede-reports/send/43-ending-drought-emergencies/4271-ede-medium-term-plan-2013-2017. Nerini, Francesco Fuso, Oliver Broad, Dimitris Mentis, Manuel Welsch, Morgan Bazilian, and Mark
Howells. 2016. “A Cost Comparison of Technology Approaches for Improving Access to Electricity Services.” Energy 95: 255–265. doi:10.1016/j.energy.2015.11.068.
Pappis, Ioannis. 2016. “Electrified Africa - Associated Investments and Costs.” Master of Science Thesis. Stockholm: KTH School of Industrial Engineering and Management. http://kth-primo.hosted.exlibrisgroup.com/primo_library/libweb/action/search.do?fn=search&ct=searc h&initialSearch=true&mode=Basic&tab=default_tab&indx=1&dum=true&srt=rank&vid=46KT H_VU1&frbg=&vl%28freeText0%29=pappis+africa&scp.scps=scope%3A%2846KTH_DATABAS ER_DS%29%2Cscope%3A%2846KTH_ALMA_DS%29%2Cscope%3A%2846KTH_DIVA_DS%29% 2Cprimo_central_multiple_fe.
Pfenninger, Stefan, and Iain Staffell. 2016. “Long-Term Patterns of European PV Output Using 30 Years of Validated Hourly Reanalysis and Satellite Data.” Energy 114: 1251–1265. doi:10.1016/j.energy.2016.08.060.
Power Africa. 2016. “Development of Kenya’s Power Sector 2015-2020.” Nairobi: USAID Kenya & East Africa.
https://www.usaid.gov/sites/default/files/documents/1860/Kenya_Power_Sector_report.pd f.
Riahi, Keywan, Arnulf Grübler, and Nebojsa Nakicenovic. 2007. “Scenarios of Long-Term Socio-Economic and Environmental Development under Climate Stabilization.” Technological
Forecasting & Social Change 74 (7): 887–935. doi:10.1016/j.techfore.2006.05.026.
Rosling, Hans. 2010. Hans Rosling and the Magic Washing Machine. Ted Talk. http://www.gapminder.org/videos/hans-rosling-and-the-magic-washing-machine/.
Sieber, Jack. 2017. “WEAP (Water Evaluation And Planning).” WEAP21. http://www.weap21.org. Staffell, Iain, and Stefan Pfenninger. 2016. “Using Bias-Corrected Reanalysis to Simulate Current and
16
Symon, Caroline. 2013. “Climate Change: Action, Trends and Implications for Business.” European Climate Foundation.
Taliotis, Constantinos, Abhishek Shivakumar, Eunice Ramos, Mark Howells, Dimitris Mentis, Vignesh Sridharan, Oliver Broad, and Linus Mofor. 2016. “An Indicative Analysis of Investment Opportunities in the African Electricity Supply Sector — Using TEMBA (The Electricity Model Base for Africa).” Energy for Sustainable Development 31 (April): 50–66. doi:10.1016/j.esd.2015.12.001.
Taylor, Karl E., Ronald J. Stouffer, and Gerald A. Meehl. 2012. “An Overview of CMIP5 and the Experiment Design.” Bulletin of the American Meteorological Society 93 (4): 485–98. doi:10.1175/BAMS-D-11-00094.1.
The Government of Kenya. 2007. “Kenya Vision 2030.” The Government of Kenya. http://www.vision2030.go.ke/home/.
The Royal Netherlands Meteorological Institute (KNMI). 2017. “Climate Explorer: KNMI Climate Change Atlas.” May 12. http://climexp.knmi.nl/plot_atlas_form.py.
The World Bank. 2012. “Access to Electricity (% of Population) | Data.” Data.worldbank.org. http://data.worldbank.org/indicator/EG.ELC.ACCS.ZS?locations=KE.
The World Bank Group. 2014. “Kenya | Data.” DataBank. The World Bank.
http://data.worldbank.org/country/kenya.
———. 2017a. “Health Nutrition and Population Statistics: Population Estimates and Projections |
DataBank.” DataBank. The World Bank.
http://databank.worldbank.org/data/reports.aspx?source=Health%20Nutrition%20and%20P opulation%20Statistics:%20Population%20estimates%20and%20projections.
———. 2017b. “World Development Indicators | DataBank.” DataBank. The World Bank. http://databank.worldbank.org/data/reports.aspx?source=world-development-indicators. The World Bank Group, and Kenya Economic Roundtable. 2017. “Kenya’s Economic Outlook to Dip in
2017.” Text/HTML. World Bank. Accessed June 5. http://www.worldbank.org/en/news/press-release/2017/04/12/kenyas-economic-outlook-to-dip-in-2017.
UNDESA. 2016. “Sustainble Development Goal 7.” Sustainable Development Knowledge Platform. https://sustainabledevelopment.un.org/sdg7.
United Nations. 2016. “Transforming Our World: The 2030 Agenda for Sustainable Development.”
Sustainable Development Knowledge Platform.
https://sustainabledevelopment.un.org/post2015/transformingourworld/publication.
USAID. 2017. “Power Africa | U.S. Agency for International Development.” March 22. https://www.usaid.gov/powerafrica.
Vuuren, Detlef P. Van van, Michel G. J. Den Elzen, Paul L. Lucas, Bas Eickhout, Bart J. Strengers, Bas Van Ruijven, Steven Wonink, and Roy Van Houdt. 2007. “Stabilizing Greenhouse Gas Concentrations at Low Levels: An Assessment of Reduction Strategies and Costs.” Climatic
Change 81 (2): 119–159.
Wahome, Mwaniki. 2013. “Kenya’s Hydro-Power Sources Drying up.” Daily Nation, January 22. http://www.nation.co.ke/lifestyle/smartcompany/Kenyas-hydro-power-sources-drying-up/1226-1670936-x965gtz/index.html.
A
APPENDIX 1 – CAPITAL COST FOR ONSSET OFF-GRID
TECHNOLOGIES
TECHNOLOGY CAPITAL COST
MINI GRID DIESEL 721 USD/kW
MINI GRID HYDRO 5000 USD/kW
MINI GRID SOLAR PHOTOVOLTAIC 4300 USD/kW
MINI GRID WIND 2500 USD/kW
STANDALONE DIESEL 938 USD/kW
STANDALONE SOLAR PHOTOVOLTAIC 5500 USD/kW
APPENDIX 2 – ENERGY CARRIERS IN THE REFERENCE ENERGY
SYSTEM
Energy carriers KEBIOMASS Biomass KECOAL Coal KEDIESEL DieselKEEL1 Electricity from power plants
KEEL2 Electricity after transmission
KEEL3I The rest of Kenya's electricity demand
KEEL3R Kenya residential electricity demand
KEOIL Oil
KEGAS Natural Gas
KESOLARPV Solar PV potential
KEURANIUM Uranium
B
APPENDIX 3 – TECHNOLOGIES IN THE REFERENCE ENERGY
SYSTEM
Technologies
KEBACKSTOP Backstop technology
KEBM00I00 Biomass import
KEBM00X00 Biomass extraction/production/refining
KEBMCHP00 Biomass cogeneration heat & power plant
KECO00I00 Coal import
KECO00X00 Coal extraction/production/refining
KECOSCP00 Coal power plant
KEDS00I00 Diesel import
KEDS00X00 Diesel extraction/production/refining
KEDSRCP03 Diesel power plants (Utility)
KEEL00T00 Transmission technology
KEEL00TDI Distribution technology (to rest of Kenya's demand)
KEEL00TDR Distribution technology (to residential demand)
KEGOVP00 Geothermal power plant
KEHF00I00 Oil import
KEHF00X00 Oil extraction technology
KEHFGCP00 Oil fired gas turbine OIL SCGT
KEHYDMS01 Hydro Power Plant (Dam) Small
KEHYDMS02 Hydro Power Plant (Dam) Medium
KEHYDMS03 Hydro Power Plant (Dam) Large
C
KEHYDMS04 Gitaru Hydro Power Plant (Dam) Large
KEHYDMS05 Masinga Hydro Power Plant (Dam) Medium
KEHYDMS06 Kamburu Hydro Power Plant (Dam) Medium
KEHYDMS07 Kindaruma Hydro Power Plant (Dam) Medium
KEHYDMS08 Kiambere Hydro Power Plant (Dam) Large
KEHYDMS09 High Grand Falls Hydro Power Plant (Dam) Large
KEHYDRRS01 Hydro Power Plant (Run-of-River)
KENG00I00 Natural Gas imported
KENG00X00 Natural Gas extraction technology
KENGCCP00 Natural Gas power plant (Combined Cycle)
KENGGCP00 Natural Gas power plant (Single Cycle)
KENULWP00 Nuclear power plant
KESOC1P00 CSP (Without storage)
KESOC2P00 CSP (With storage)
KESOC3P00 CSP (With gas firing)
KESOPV Solar potential
KESOU1P3 Solar Photovoltaic (Utility)
KEUR00I00 Uranium import
KEUR00X00 Uranium extraction/production/refining
KEWI25P00 Wind (Offshore, 17% Capacity Factor)
KEWI30P00 Wind (Onshore, 43% Capacity Factor)
MG DIESEL Mini Grid Diesel
MG HYDRO Mini Grid Hydro
D
MG PHOTOVOLTIC Mini Grid Photovoltaic
MG WIND Mini Grid Wind
NATIONAL GRID National Grid
SA DIESEL Standalone Diesel
SA PHOTOVOLTIC Standalone Photovoltaic
E
APPENDIX 4 – DATA COLLECTION OSEMOSYS MODEL
APPENDIX 4.1 – TIMESLICES OSEMOSYS
TIMESLICE SHARE OF YEAR
JAND 0.0425 JANE 0.0142 JANN 0.0283 FEBD 0.0384 FEBE 0.0128 FEBN 0.0256 MARD 0.0425 MARE 0.0142 MARN 0.0283 APRD 0.0411 APRE 0.0137 APRN 0.0274 MAYD 0.0425 MAYE 0.0142 MAYN 0.0283 JUND 0.0411 JUNE 0.0137 JUNN 0.0274 JULD 0.0425 JULE 0.0142 JULN 0.0283 AUGD 0.0425 AUDE 0.0142 AUGN 0.0283 SEPD 0.0411 SEPE 0.0137 SEPN 0.0274 OCTD 0.0425 OCTE 0.0142 OCTN 0.0283 NOVD 0.0411 NOVE 0.0137 NOVN 0.0274 DECD 0.0425 DECE 0.0142 DECN 0.0283
F
APPENDIX 4.2 – TECHNOLOGY COSTS
CAPITAL COST (MUSD/GW) FIXED COST (MUSD/GW) VARIABLE COST (MUSD/PJ)
2015 2030 2015 2030 2015 2030 KEBACKSTOP 9999 9999 9999 9999 9999 9999 KEBM00I00 0 0 0 0 1.5 1.5 KEBM00X00 0 0 0 0 1.5 1.5 KEBMCHP00 5094.98099 4836.855152 57.66588012 54.95856183 0.7977776741 0.7603233231 KECO00I00 0 0 0 0 3.14 5.185 KECO00X00 0 0 0 0 2.81 4.85 KECOSCP00 2528.253 2528.253 26.74815973 26.74815973 1.160944433 1.160944433 KEDS00I00 0 0 0 0 13.7 27.375 KEDS00X00 0 0 0 0 13.7 27.375 KEDSRCP03 780.57 780.57 8 8 0.5555555556 0.5555555556 KEEL00T00 0 0 0 0 0 0 KEEL00TDI 0 0 0 0 0 0 KEEL00TDR 0 0 0 0 0 0 KEGOCVP00 5894.593096 5462.089624 48.625 45.33333333 0 0 KEHF00I00 0 0 0 0 7.935 15.9 KEHF00X00 0 0 0 0 7.935 15.9 KEHFGCP00 1488.375 1488.375 0 0 4.166666667 4.166666667 KENG00I00 0 0 0 0 12.32 15.6 KENG00X00 0 0 0 0 7.2 10.2 KENGCCP00 1181.935125 1181.935125 1.323058618 1.323058618 0.6595155211 0.6595155211 KENGGCP00 780.57 780.57 8.328946435 8.328946435 0.5576278211 0.5576278211 KESOC1P00 5189.039185 3558.528283 179.90625 125.0833333 0 0 KESOC2P00 6654.65323 3878.753357 54.06503532 41.7775273 0.6007226145 0.4641947472 KESOC3P00 1590.57675 1590.57675 0 0 4.555555556 4.555555556 KESOPV 0 0 0 0 0 0 KESOU1P03 1517.748327 1085.822567 24.25 22.33333333 0 0 KEWI25P00 2126.168903 1922.652502 76.1 75.339 0 0 KEWI30P00 2126.168903 1922.652502 76.1 75.339 0 0 KEHYDMS01 3000 3000 27 27 0.13988889 0.13988889 KEHYDMS02 3933.2 3933.2 17.5 17.5 0.13988889 0.13988889 KEHYDMS03 3325.533 3325.533 27.58274 27.58274 0.13988889 0.13988889 KEHYRRS01 3430 3430 27.4 27.4 0.13988889 0.13988889 KEHYDMS04 - - 27.4 27.4 0.13988889 0.13988889 KEHYDMS05 - - 27.4 27.4 0.13988889 0.13988889 KEHYDMS06 - - 27.4 27.4 0.13988889 0.13988889 KEHYDMS07 - - 27.4 27.4 0.13988889 0.13988889 KEHYDMS08 - - 27.4 27.4 0.13988889 0.13988889 KEHYDMS09 3708 3708 16 16 0.13888889 0.13888889 KENULWP00 8068 8068 7.5 7.5 2.77777778 2.77777778 KEUR00I00 0 0 0 0 2.8 2.8 KEUR00X00 0 0 0 0 2.8 2.8
G
APPENDIX 4.3 – TECHNOLOGY PERFORMANCE
INPUT ACTIVITY RATIO OUTPUT ACTIVITY RATIO AVAILABILITY FACTOR
CAPACITY FACTOR OPERATIONAL LIFE (YEARS) KEBACKSTOP 1 1 1 KEBM00I00 1 1 35 KEBM00X00 1 1 35 KEBMCHP00 2.631578 1 0.93 0.53 30 KECO00I00 1 1 35 KECO00X00 1 0 35 KECOSCP00 2.7027 1 0.94 0.85 40 KEDS00I00 1 1 30 KEDS00X00 1 1 30 KEDSRCP03 2.85714 1 0.9 0.9 25 KEEL00T00 1 0.959 1 60 KEEL00TDI 1 0.8603 1 60 KEEL00TDR 1 0.8603 1 60 KEGOCVP00 1 0.914 0.93 25 KEHF00I00 1 1 25 KEHF00X00 1 1 25 KEHFGCP00 2.85714 1 0.9 0.9 25 KENG00I00 1 1 100 KENG00X00 1 1 30 KENGCCP00 2.083333 1 0.93 0.935 30 KENGGCP00 3.33 1 0.93 0.935 25 KESOC1P00 1 1 0.7 / 0 / 0 * 25 KESOC2P00 1 1 0.71 / 0.55 / 0.55 * 25 KESOC3P00 1.886792 1 0.93 0.92 25 KESOPV 1 1 100 KESOU1P03 1 1 1 0.5 / 0 / 0 * 25
KEWI25P00 1 0.9 See Appendix 4.4 25
KEWI30P00 1 0.85 See Appendix 4.4 25
KEHYDMS01 1 0.944 0.45 40
KEHYDMS02 1 0.944 Based on results from WEAP 50
KEHYDMS03 1 0.944 Based on results from WEAP 50
KEHYRRS01 1 0.944 0.512 50
KEHYDMS04 1 0.944 Based on results from WEAP 50
KEHYDMS05 1 0.944 Based on results from WEAP 50
KEHYDMS06 1 0.944 Based on results from WEAP 50
KEHYDMS07 1 0.944 Based on results from WEAP 50
KEHYDMS08 1 0.944 Based on results from WEAP 50
KEHYDMS09 1 0.944 - 50
KENULWP00 2.85714 1 0.9 0.85 40
KEUR00I00 1 1 25
KEUR00X00 1 1 25
* Day / Evening / Night
H
APPENDIX 4.4 – CAPACITY FACTORS WIND
The Capacity Factors for Wind Power Plants were extracted using the online tool Renewable Ninja, accessed at https://www.renewables.ninja/# (Staffell and Pfenninger 2016; Pfenninger and Staffell 2016). The position for accessing wind capacity factors for KEWI25P00 was Meru Wind Farm (planned) at (latitude 0.184, longitude 36.631) and for KEWI30P00 at Lake Turkana Wind Farm (under construction) at (latitude 2.8, longitude 36.832). An assumption was made that wind capacity factors follow the same patterns during the entire modelling period.
TIMESLICE KEWI25P00 KEWI30P00
JAND 0.115 0.449 JANE 0.115 0.449 JANN 0.115 0.449 FEBD 0.102 0.327 FEBE 0.102 0.327 FEBN 0.102 0.327 MARD 0.166 0.452 MARE 0.166 0.452 MARN 0.166 0.452 APRD 0.197 0.457 APRE 0.197 0.457 APRN 0.197 0.457 MAYD 0.244 0.512 MAYE 0.244 0.512 MAYN 0.244 0.512 JUND 0.175 0.485 JUNE 0.175 0.485 JUNN 0.175 0.485 JULD 0.15 0.417 JULE 0.15 0.417 JULN 0.15 0.417 AUGD 0.178 0.401 AUDE 0.178 0.401 AUGN 0.178 0.401 SEPD 0.226 0.524 SEPE 0.226 0.524 SEPN 0.226 0.524 OCTD 0.222 0.424 OCTE 0.222 0.424 OCTN 0.222 0.424 NOVD 0.158 0.388 NOVE 0.158 0.388 NOVN 0.158 0.388 DECD 0.134 0.392 DECE 0.134 0.392 DECN 0.134 0.392
